ICLR 2022 论文列表

We propose a new method for spatio-temporal forecasting on arbitrarily distributed points. Assuming that the observed system follows an unknown partial differential equation, we derive a continuous-time model for the dynamics of the data via the finite element method. The resulting graph neural network estimates the instantaneous effects of the unknown dynamics on each cell in a meshing of the spatial domain. Our model can incorporate prior knowledge via assumptions on the form of the unknown PDE, which induce a structural bias towards learning specific processes. Through this mechanism, we derive a transport variant of our model from the convection equation and show that it improves the transfer performance to higher-resolution meshes on sea surface temperature and gas flow forecasting against baseline models representing a selection of spatio-temporal forecasting methods. A qualitative analysis shows that our model disentangles the data dynamics into their constituent parts, which makes it uniquely interpretable.

Self-supervised visual representation learning has recently attracted significant research interest. While a common way to evaluate self-supervised representations is through transfer to various downstream tasks, we instead investigate the problem of measuring their interpretability, i.e. understanding the semantics encoded in raw representations. We formulate the latter as estimating the mutual information between the representation and a space of manually labelled concepts. To quantify this we introduce a decoding bottleneck: information must be captured by simple predictors, mapping concepts to clusters in representation space. This approach, which we call reverse linear probing, provides a single number sensitive to the semanticity of the representation. This measure is also able to detect when the representation contains combinations of concepts (e.g., "red apple'') instead of just individual attributes ("red'' and "apple'' independently). Finally, we propose to use supervised classifiers to automatically label large datasets in order to enrich the space of concepts used for probing. We use our method to evaluate a large number of self-supervised representations, ranking them by interpretability, highlight the differences that emerge compared to the standard evaluation with linear probes and discuss several qualitative insights. Code at: https://github.com/iro-cp/ssl-qrp.

We present a novel adaptive optimization algorithm for large-scale machine learning problems. Equipped with a low-cost estimate of local curvature and Lipschitz smoothness, our method dynamically adapts the search direction and step-size. The search direction contains gradient information preconditioned by a well-scaled diagonal preconditioning matrix that captures the local curvature information. Our methodology does not require the tedious task of learning rate tuning, as the learning rate is updated automatically without adding an extra hyper-parameter. We provide convergence guarantees on a comprehensive collection of optimization problems, including convex, strongly convex, and nonconvex problems, in both deterministic and stochastic regimes. We also conduct an extensive empirical evaluation on standard machine learning problems, justifying our algorithm's versatility and demonstrating its strong performance compared to other start-of-the-art first-order and second-order methods.

Existing approaches to lifelong language learning rely on plenty of labeled data for learning a new task, which is hard to obtain in most real scenarios. Considering that humans can continually learn new tasks from a handful of examples, we expect the models also to be able to generalize well on new few-shot tasks without forgetting the previous ones. In this work, we define this more challenging yet practical problem as Lifelong Few-shot Language Learning (LFLL) and propose a unified framework for it based on prompt tuning of T5. Our framework called LFPT5 takes full advantage of PT's strong few-shot learning ability, and simultaneously trains the model as a task solver and a data generator. Before learning a new domain of the same task type, LFPT5 generates pseudo (labeled) samples of previously learned domains, and later gets trained on those samples to alleviate forgetting of previous knowledge as it learns the new domain. In addition, a KL divergence loss is minimized to achieve label consistency between the previous and the current model. While adapting to a new task type, LFPT5 includes and tunes additional prompt embeddings for the new task. With extensive experiments, we demonstrate that LFPT5 can be applied to various different types of tasks and significantly outperform previous methods in different LFLL settings.

Inferring interactions among entities is an important problem in studying dynamical systems, which greatly impacts the performance of downstream tasks, such as prediction. In this paper, we tackle the relational inference problem in a setting where each entity can potentially have a set of individualized information that other entities cannot have access to. Specifically, we represent the system using a graph in which the individualized information become node-specific information (NSI). We build our model in the framework of Neural Relation Inference (NRI), where the interaction among entities are uncovered using variational inference. We adopt NRI model to incorporate the individualized information by introducing private nodes in the graph that represent NSI. Such representation enables us to uncover more accurate relations among the agents and therefore leads to better performance on the downstream tasks. Our experiment results over real-world datasets validate the merit of our proposed algorithm.

This work targets automated designing and scaling of Vision Transformers (ViTs). The motivation comes from two pain spots: 1) the lack of efficient and principled methods for designing and scaling ViTs; 2) the tremendous computational cost of training ViT that is much heavier than its convolution counterpart. To tackle these issues, we propose As-ViT, an auto-scaling framework for ViTs without training, which automatically discovers and scales up ViTs in an efficient and principled manner. Specifically, we first design a "seed" ViT topology by leveraging a training-free search process. This extremely fast search is fulfilled by a comprehensive study of ViT's network complexity, yielding a strong Kendall-tau correlation with ground-truth accuracies. Second, starting from the "seed" topology, we automate the scaling rule for ViTs by growing widths/depths to different ViT layers. This results in a series of architectures with different numbers of parameters in a single run. Finally, based on the observation that ViTs can tolerate coarse tokenization in early training stages, we propose a progressive tokenization strategy to train ViTs faster and cheaper. As a unified framework, As-ViT achieves strong performance on classification (83.5% top1 on ImageNet-1k) and detection (52.7% mAP on COCO) without any manual crafting nor scaling of ViT architectures: the end-to-end model design and scaling process costs only 12 hours on one V100 GPU. Our code is available at https://github.com/VITA-Group/AsViT.

This paper introduces Multi-Agent MDP Homomorphic Networks, a class of networks that allows distributed execution using only local information, yet is able to share experience between global symmetries in the joint state-action space of cooperative multi-agent systems. In cooperative multi-agent systems, complex symmetries arise between different configurations of the agents and their local observations. For example, consider a group of agents navigating: rotating the state globally results in a permutation of the optimal joint policy. Existing work on symmetries in single agent reinforcement learning can only be generalized to the fully centralized setting, because such approaches rely on the global symmetry in the full state-action spaces, and these can result in correspondences across agents. To encode such symmetries while still allowing distributed execution we propose a factorization that decomposes global symmetries into local transformations. Our proposed factorization allows for distributing the computation that enforces global symmetries over local agents and local interactions. We introduce a multi-agent equivariant policy network based on this factorization. We show empirically on symmetric multi-agent problems that globally symmetric distributable policies improve data efficiency compared to non-equivariant baselines.

Rate-distortion (R-D) function, a key quantity in information theory, characterizes the fundamental limit of how much a data source can be compressed subject to a fidelity criterion, by any compression algorithm. As researchers push for ever-improving compression performance, establishing the R-D function of a given data source is not only of scientific interest, but also reveals the possible room for improvement in existing compression algorithms. Previous work on this problem relied on distributional assumptions on the data source (Gibson, 2017) or only applied to discrete data (Blahut, 1972; Arimoto, 1972). By contrast, this paper makes the first attempt at an algorithm for sandwiching the R-D function of a general (not necessarily discrete) source requiring only i.i.d. data samples. We estimate R-D sandwich bounds for a variety of artificial and real-world data sources, in settings far beyond the feasibility of any known method, and shed light on the optimality of neural data compression (Ballé et al., 2021; Yang et al., 2022). Our R-D upper bound on natural images indicates theoretical room for improving state-of-the-art image compression methods by at least one dB in PSNR at various bitrates. Our data and code can be found at https://github.com/mandt-lab/RD-sandwich.

Communication is a key bottleneck in federated learning where a large number of edge devices collaboratively learn a model under the orchestration of a central server without sharing their own training data. While local SGD has been proposed to reduce the number of FL rounds and become the algorithm of choice for FL, its total communication cost is still prohibitive when each device needs to communicate with the remote server repeatedly for many times over bandwidth-limited networks. In light of both device-to-device (D2D) and device-to-server (D2S) cooperation opportunities in modern communication networks, this paper proposes a new federated optimization algorithm dubbed hybrid local SGD (HL-SGD) in FL settings where devices are grouped into a set of disjoint clusters with high D2D communication bandwidth. HL-SGD subsumes previous proposed algorithms such as local SGD and gossip SGD and enables us to strike the best balance between model accuracy and runtime. We analyze the convergence of HL-SGD in the presence of heterogeneous data for general nonconvex settings. We also perform extensive experiments and show that the use of hybrid model aggregation via D2D and D2S communications in HL-SGD can largely speed up the training time of federated learning.

This paper considers few-shot learning under the cross-domain scenario. The cross-domain setting imposes a critical challenge, i.e., using very few (support) samples to generalize the already-learned model to a novel domain. We hold a hypothesis, i.e., if a deep model is capable to fast generalize itself to different domains (using very few samples) during training, it will maintain such domain generalization capacity for testing. It motivates us to propose a novel Domain-Switch Learning (DSL) framework. DSL embeds the cross-domain scenario into the training stage in a ``fast switching'' manner. Specifically, DSL uses a single domain for a training iteration and switches into another domain for the following iteration. During the switching, DSL enforces two constraints: 1) the deep model should not over-fit the domain in the current iteration and 2) the deep model should not forget the already-learned knowledge of other domains. These two constraints jointly promote fast generalization across different domains. Experimental results confirm that the cross-domain generalization capacity can be inherited from the training stage to the testing stage, validating our key hypothesis. Consequentially, DSL significantly improves cross-domain few-shot classification and sets up new state of the art.

In this work we study the MARINA method of Gorbunov et al (ICML, 2021) -- the current state-of-the-art distributed non-convex optimization method in terms of theoretical communication complexity. Theoretical superiority of this method can be largely attributed to two sources: a carefully engineered biased stochastic gradient estimator, which leads to a reduction in the number of communication rounds, and the reliance on {\em independent} stochastic communication compression, which leads to a reduction in the number of transmitted bits within each communication round. In this paper we i) extend the theory of MARINA to support a much wider class of potentially {\em correlated} compressors, extending the reach of the method beyond the classical independent compressors setting, ii) show that a new quantity, for which we coin the name {\em Hessian variance}, allows us to significantly refine the original analysis of MARINA without any additional assumptions, and iii) identify a special class of correlated compressors based on the idea of {\em random permutations}, for which we coin the term Perm$K$, the use of which leads to up to $O(\sqrt{n})$ (resp. $O(1 + d/\sqrt{n})$) improvement in the theoretical communication complexity of MARINA in the low Hessian variance regime when $d\geq n$ (resp. $d \leq n$), where $n$ is the number of workers and $d$ is the number of parameters describing the model we are learning. We corroborate our theoretical results with carefully engineered synthetic experiments with minimizing the average of nonconvex quadratics, and on autoencoder training with the MNIST dataset.

To attain higher efficiency, the industry has gradually reformed towards application-specific hardware accelerators. While such a paradigm shift is already starting to show promising results, designers need to spend considerable manual effort and perform large number of time-consuming simulations to find accelerators that can accelerate multiple target applications while obeying design constraints. Moreover, such a simulation-driven approach must be re-run from scratch every time the set of target applications or design constraints change. An alternative paradigm is to use a data-driven, offline approach that utilizes logged simulation data, to architect hardware accelerators, without needing any form of simulations. Such an approach not only alleviates the need to run time-consuming simulation, but also enables data reuse and applies even when set of target applications changes. In this paper, we develop such a data-driven offline optimization method for designing hardware accelerators, dubbed PRIME, that enjoys all of these properties. Our approach learns a conservative, robust estimate of the desired cost function, utilizes infeasible points and optimizes the design against this estimate without any additional simulator queries during optimization. PRIME architects accelerators---tailored towards both single- and multi-applications---improving performance upon stat-of-the-art simulation-driven methods by about 1.54x and 1.20x, while considerably reducing the required total simulation time by 93% and 99%, respectively. In addition, PRIME also architects effective accelerators for unseen applications in a zero-shot setting, outperforming simulation-based methods by 1.26x.

Catastrophic forgetting presents a challenge in developing deep learning models capable of continual learning, i.e. learning tasks sequentially. Recently, both computer vision and natural-language processing have witnessed great progress through the use of large-scale pretrained models. In this work, we present an empirical study of catastrophic forgetting in this pretraining paradigm. Our experiments indicate that large, pretrained ResNets and Transformers are significantly more resistant to forgetting than randomly-initialized, trained-from-scratch models; this robustness systematically improves with scale of both model and pretraining dataset size. We take initial steps towards characterizing what aspect of model representations allows them to perform continual learning so well, finding that in the pretrained models, distinct class representations grow more orthogonal with scale. Our results suggest that, when possible, scale and a diverse pretraining dataset can be useful ingredients in mitigating catastrophic forgetting.

We introduce DictFormer with the efficient shared dictionary to provide a compact, fast, and accurate transformer model. DictFormer significantly reduces the redundancy in the transformer's parameters by replacing the prior transformer's parameters with a compact, shared dictionary, few unshared coefficients, and indices. Also, DictFormer enables faster computations since expensive weights multiplications are converted into cheap shared look-ups on dictionary and few linear projections. Training dictionary and coefficients are not trivial since indices used for looking up dictionary are not differentiable. We adopt a sparse-constraint training with $l_1\,\,norm$ relaxation to learn coefficients and indices in DictFormer. DictFormer is flexible to support different model sizes by dynamically changing dictionary size. Compared to existing lightweight Transformers, DictFormer consistently reduces model size over Transformer on multiple tasks, e.g., machine translation, abstractive summarization, and language modeling. Extensive experiments show that DictFormer reduces $3.6\times$ to $8.9\times$ model size with similar accuracy over multiple tasks, compared to Transformer.

With recent progress in joint modeling of visual and textual representations, Vision-Language Pretraining (VLP) has achieved impressive performance on many multimodal downstream tasks. However, the requirement for expensive annotations including clean image captions and regional labels limits the scalability of existing approaches, and complicates the pretraining procedure with the introduction of multiple dataset-specific objectives. In this work, we relax these constraints and present a minimalist pretraining framework, named Simple Visual Language Model (SimVLM). Unlike prior work, SimVLM reduces the training complexity by exploiting large-scale weak supervision, and is trained end-to-end with a single prefix language modeling objective. Without utilizing extra data or task-specific customization, the resulting model significantly outperforms previous pretraining methods and achieves new state-of-the-art results on a wide range of discriminative and generative vision-language benchmarks, including VQA (+3.74% vqa-score), NLVR2 (+1.17% accuracy), SNLI-VE (+1.37% accuracy) and image captioning tasks (+10.1% average CIDEr score). Furthermore, we demonstrate that SimVLM acquires strong generalization and transfer ability, enabling zero-shot behavior including open-ended visual question answering and cross-modality transfer.

Protein complex formation is a central problem in biology, being involved in most of the cell's processes, and essential for applications, e.g. drug design or protein engineering. We tackle rigid body protein-protein docking, i.e., computationally predicting the 3D structure of a protein-protein complex from the individual unbound structures, assuming no conformational change within the proteins happens during binding. We design a novel pairwise-independent SE(3)-equivariant graph matching network to predict the rotation and translation to place one of the proteins at the right docked position relative to the second protein. We mathematically guarantee a basic principle: the predicted complex is always identical regardless of the initial locations and orientations of the two structures. Our model, named EquiDock, approximates the binding pockets and predicts the docking poses using keypoint matching and alignment, achieved through optimal transport and a differentiable Kabsch algorithm. Empirically, we achieve significant running time improvements and often outperform existing docking software despite not relying on heavy candidate sampling, structure refinement, or templates.

Numerous recent works utilize bi-Lipschitz regularization of neural network layers to preserve relative distances between data instances in the feature spaces of each layer. This distance sensitivity with respect to the data aids in tasks such as uncertainty calibration and out-of-distribution (OOD) detection. In previous works, features extracted with a distance sensitive model are used to construct feature covariance matrices which are used in deterministic uncertainty estimation or OOD detection. However, in cases where there is a distribution over tasks, these methods result in covariances which are sub-optimal, as they may not leverage all of the meta information which can be shared among tasks. With the use of an attentive set encoder, we propose to meta learn either diagonal or diagonal plus low-rank factors to efficiently construct task specific covariance matrices. Additionally, we propose an inference procedure which utilizes scaled energy to achieve a final predictive distribution which is well calibrated under a distributional dataset shift.

Traditional computer vision models are trained to predict a fixed set of predefined categories. Recently, natural language has been shown to be a broader and richer source of supervision that provides finer descriptions to visual concepts than supervised "gold" labels. Previous works, such as CLIP, use InfoNCE loss to train a model to predict the pairing between images and text captions. CLIP, however, is data hungry and requires more than 400M image-text pairs for training. The inefficiency can be \textit{partially} attributed to the fact that the image-text pairs are noisy. To address this, we propose OTTER (Optimal TransporT distillation for Efficient zero-shot Recognition), which uses online entropic optimal transport to find a soft image-text match as labels for contrastive learning. Based on pretrained image and text encoders, models trained with OTTER achieve strong performance with only 3M image text pairs. Compared with InfoNCE loss, label smoothing, and knowledge distillation, OTTER consistently outperforms these baselines in zero-shot evaluation on Google Open Images (19,958 classes) and multi-labeled ImageNet 10K (10032 classes) from Tencent ML-Images. Over 42 evaluations on 7 different dataset/architecture settings x 6 metrics, OTTER outperforms (32) or ties (2) all baselines in 34 of them. Our source code is open sourced at https://github.com/facebookresearch/OTTER.

Learning in games has become an object of intense interest for ML due to its connections to numerous AI architectures. We study standard online learning in games but from a non-standard perspective. Instead of studying the behavior of a single initial condition and whether it converges to equilibrium or not, we study the behavior of a probability distribution/measure over a set of initial conditions. This initial uncertainty is well-motivated both from a standard game-theoretic perspective (e.g. a modeler's uncertainty about the agents' initial beliefs) as well as from a ML one (e.g. noisy measurements, system initialization from a dataset distribution). Despite this, little is formally known about whether and under what conditions uncertainty is amplified or reduced in these systems. We use the popular measure of differential entropy to quantify the evolution of uncertainty. We find that such analysis shares an intimate relationship with volume analysis, a technique which was recently used to demonstrate the occurrence of Lyapunov chaos when using Multiplicative Weights Update (MWU) or Follow-the-Regularized-Leader (FTRL) algorithms in zero-sum games. This allows us to show that the differential entropy of these learning-in-game systems increases linearly with time, formalizing their increased unpredictability over time. We showcase the power of the framework by applying it in the study of multiple related systems, including different standard online optimization algorithms in numerous games and dynamics of evolutionary game theory.